How much energy does it take to print a metal part, and where does that energy actually go?

When discussing the energy consumption of a metal 3D printer, the laser is almost always the main expense. And the laser is, by far, the smallest part of the overall cost.

It is important to understand the actual distribution, because it determines what is worth optimizing and because it allows for an honest comparison with other technologies.

Where does the energy go?

Consumer	Weight on the Bill	Why
Platform Heating	High	Keeping the platform warm for tens of hours at a time consumes more energy than the laser itself
Inert Gas Circuit	High	The blower that maintains laminar flow over the bed runs continuously throughout the entire build
Filtration	Medium	Removing fumes and condensate from the gas circuit
Laser and Optics	Medium-Low	A fiber laser has good electrical efficiency and is only on while melting: a fraction of the cycle
Hall Cooling	Medium	The machine heats the hall, and the hall needs to be cooled
Compressed Air	Medium	It's almost never accounted for and is almost always present
Post-Processing Ovens	Very High	Stress relief and heat treatments involve hours of oven time at temperature. If there's hot isostatic pressing, even more

The latter is what breaks all comparisons that only measure the machine: post-processing can consume as much or more than build, and it is systematically ignored.

And one item that isn't electrical but is still significant: argon. Producing and transporting it has its own footprint, and a long build consumes gas continuously, especially during the initial purging.

What really reduces consumption per part

It's not about changing machines. It's about filling the build plate.

The heating, blower, filtration, and air conditioning consume virtually the same amount of energy whether the build plate contains one part or thirty. All of that energy is distributed among the parts being processed. A build plate that is only 20% full uses several times the energy per part as the same build plate when full.

From there come the four royal levers, in order of effect:

Optimize the build plate. Group orders, print multiple units simultaneously, and combine parts from different projects using the same alloy.
Reduce the build height. Fewer layers mean less printing time per hour. Orientation is an energy-related decision as well as a dimensional one.
Reduce supports. This reduces the amount of molten material wasted, laser time, and operator hours spent removing supports. (See supports.
Group oven cycles. Process multiple builds in the same batch, instead of firing the oven for each one.

None of the four require purchasing anything. All four depend on how the work is planned.

An honest comparison with machining

This is where almost all studies cheat, in one way or another. For the comparison to be meaningful, the same values must be counted in both columns.

To be taken into account	In additive manufacturing	In machining
Energy to manufacture the starting material	Powder atomization, which is not energy-inefficient	Casting and rolling of the raw material
Material that does not end up in the part	Supports, and powder removed due to degradation	The chips, which in some parts are 80-90% of the raw material
Process energy	Complete machine, not just the laser	Machine, tool, coolant
Post-processing	Furnaces, and it is usually a significant amount	Treatments, if any
Rejected parts	In both columns	In both columns

All things considered, the result isn't a headline but a rule: additive manufacturing is more energy-efficient when the part is complex and machining would discard most of the raw material, and less efficient when the part is simple and comes from a small raw material. This is the same boundary that separates when it's more cost-effective, and it lies in printing or machining.

Anyone who tells you that additive manufacturing is always more energy efficient is selling you a line. Anyone who tells you it never is, too.

What you should measure in your own plant

If you really want to know and not just estimate, three numbers are enough to get started:

kWh per build, with a meter on the machine line, not the building line.
Average build plate utilization, as a percentage. This is the number with the most room for improvement, and almost no one tracks it.
kWh of oven per batch, counting how many builds go into each batch.

These three figures are used to calculate the consumption per good part, which is the only figure useful for comparison and improvement. Furthermore, this is the type of data that will soon be required in writing: it is included in is 3D metal printing sustainable? and in European ecodesign.

The actual consumption depends mainly on how many machine hours each good part takes, so the starting figures are in the data sheets for ALBA 300 and ALBA 500.